For the complete documentation index, see llms.txt. This page is also available as Markdown.

AWS

The procedure to set up CCM for AWS by using Harness Self-Managed Enterprise Edition.

This topic walks you through the steps required to set up CACM for AWS in a self-managed platform.

Figure: AWS CACM Self-Managed Enterprise Edition architecture diagram

Click to view full size image

You need to perform the following tasks to set up CACM for AWS:

Add a new user for programmatic access

  1. Sign in to your AWS console.

  2. Create a new user. For example, harness-ccm-service-user. For more information, go to Adding a user.

  3. Set up programmatic user access keys by using the policy given below.

Policy to set up programmatic user access

Make a note of your AWS Access key and Secret key.

Create Amazon S3 Bucket

  1. Create an S3 bucket with the following naming convention. For more information, go to Creating a bucket. This bucket will be used to retrieve the CUR report from your master AWS accounts.

harness-ccm-service-data-bucket-<accountid>

  1. Create another bucket with the following naming convention.

harness-ccm-service-template-bucket-<accountid>

  1. Apply the following bucket policy.

Edit the CF template

In the given YAML template, modify the specified field.

Replace <accountid> with the AWS account ID in which the S3 buckets were created.

YAML template

Upload the CF template to S3 bucket

Upload the YAML template to S3 bucket harness-ccm-service-template-bucket-<accountid>.

  1. Go to Amamzon S3 > ccm-service-template-bucket.

  2. Create a folder named v1.

  3. Upload the file.

Click to view full size image

Make a note of the following:

  • The names of the two S3 buckets.

  • The complete path of the CF template in the bucket. For example, https://ccm-service-template-bucket.s3.amazonaws.com/v1/HarnessCCMServiceAWSTemplate.yaml.

Deploy workloads via Helm charts

  1. Clone the chart repository.

  2. Upgrade charts if you're already using Harness Self-managed Enterprise Edition services. Perform the following steps to update the override files:

    1. Retrieve the current override values provided during the installation or upgrade of the Helm charts.

    2. After obtaining the override file, you can make necessary modifications based on the type of environment.

Override file changes for a connected environment
Override file changes for an air-gapped environment

CACM leverages AWS APIs that require connectivity from the isolated (air-gapped) instance. To grant access to these AWS APIs, establish VPC endpoints for the respective AWS services. For services lacking VPC endpoints, use a proxy to facilitate access.

  1. After making the necessary updates to the override file, you can proceed with the Helm chart upgrade.

Handling Kubernetes secrets

When installing or upgrading the Helm charts, Kubernetes secrets with default values are created within the cluster. These generated secrets should be updated with the values mentioned above. Before updating the secrets, you need to convert the secret into base64 encoded format. For example, if your AWS_DESTINATION_BUCKET value is "harness-ccm-service-data-bucket-12345678", it would be stored as aGFybmVzcy1jY20tc2VydmljZS1kYXRhLWJ1Y2tldC0xMjM0NTY3OA== after encoding. After changing secrets, we will provide directives to kubectl delete the corresponding pods in order for your release to inherit new changes.

The following are the secrets specific to CACM services:

  • batch-processing

  • cloud-info-secret-mount [config-file]

  • nextgen-ce

Config file

The following are some secrets from platform-service that you need to update:

  • smtp-secret - Required to support budget alerts email.

To increase TimescaleDB to 100Gi, run: kubectl edit pvc wal-volume-harness-timescaledb-0 -n <namespace>. Features like Recommendations and Anomalies within CACM services use it.

Next steps

Troubleshooting

If in case the K8s secrets expire, the secrets will have to be set again. First you would have to update the secrets in respective secret.yaml and then delete the pod. We recommend to kubectl delete the following pods:

  • batch-processing

  • ce-nextgen

  • cloud-info

and then follow the same steps to set the keys. After the new keys are set, verify the changes by looking at the configs for the pods.

Please refer to the commands below:

  • batch-processing: kubectl exec -it -n <namespace> batch-processing cat batch-processing-config.yml | grep -E 'awsAccessKey|awsSecretKey

  • cloud-info:

kubectl exec -it -n <namespace> cloud-info cat config/config.toml | grep -E 'accessKey|secretKey'

  • nextgen-ce: kubectl exec -it -n <namespace> nextgen-ce cat config.yml | grep -E 'accessKey|secretKey|harnessAwsAccountId|destinationBucket:|awsConnectorTemplate'

AWS Visibility connector connection test fails with implicit deny on organizations:ListAccounts or organizations:ListTagsForResource in SMP

Connectors created after April 8, 2026 require organizations:ListAccounts and organizations:ListTagsForResource on the IAM role. Re-run your CloudFormation stack with the updated YAML template to attach the updated HarnessEventsMonitoringPolicy, then retest the connector.

Last updated

Was this helpful?